Blog

AI agents, RAG, and integrations — notes from production.

Sizing and Staffing the Human Review Queue for Ambiguous Detections

The computer vision human review queue takes the ambiguous confidence band - size it by detections times band width, and keep each review to seconds.

MQTT Quality of Service: Getting a Detection Event Out of a Flaky Store

MQTT QoS for retail edge: use QoS 1 (at-least-once) plus downstream dedupe for stockout alerts - QoS 0 can drop them, QoS 2 is overkill for idempotent events.

Fixed Cameras vs Staff-Phone Capture for Shelf Monitoring

Shelf monitoring camera setup: fixed cameras give consistent input and a tighter model; staff-phone capture is near-free to roll out but adds accuracy variance.

Confidence Calibration for a Detection Model: Making the Score Mean Something

Detection model confidence calibration: a raw 0.9 isn't 90% accurate, and confidence routing sets thresholds on that number. Fix it with temperature scaling.

Model Quantization for Edge Inference: Trading a Little Accuracy for Speed

Model quantization for edge inference converts 32-bit weights to fp16 or int8 - fp16 is near-lossless, int8 needs calibration and a measured accuracy check.

ONNX Runtime and Model Portability Across Mixed Store Hardware

ONNX Runtime for edge deployment: export the model once, run it through whichever execution provider matches each store's hardware - no per-device build.

Single-Shot vs Two-Stage vs Transformer Detectors: Why YOLO Wins on the Edge

An object detection model for a retail shelf runs on edge hardware, so single-shot detectors win on speed - two-stage and transformer models are slower.

Search Analytics: The Queries That Tell You What's Wrong With Your Catalog

Ecommerce search analytics is honest customer feedback - zero-results, high-exit, and low-conversion queries each point at a catalog or relevance problem.

Zero-Results Pages: Turning a Dead End Into a Recovery

A zero results search page is a near-certain exit - recover it with labelled relaxed results, a did-you-mean, nearest categories, and by capturing the query.

Store Photos and Privacy: Handling Staff and Customers in Shelf Images

Retail computer vision privacy: shelf photos catch people incidentally. Process on-device, keep only the result, retain no images, add no face recognition.